Remotely sensed identification of canopy characteristics using UAV-based imagery under unstable environmental conditions

Water is a crucial element for plant growth, metabolic processes, and general health. Water-deficit, typically simplified by drought stress, is the most critical photosynthetic source of stress that restricts plant growth, crop yield, and food product quality. This research highlights state of the art, possibilities for detecting the canopy temperature by integrating very-high-resolution RGB and thermal imagery from UAV. A multi-rotor drone has assembled by DJI (S900) attached with RGB thermal and cameras was used for experiments. The thermal cameras have a spectral range of 7.5–13μm, a resolution of 640 × 512 pixels, thermal sensitivity of <0.05 °C at ＋30 °C, and a focal length of 25 mm, respectively. UAV flights were operated with DJI ground stations pro software (SZ DJI Technology Co. Ltd., China), using the DJI A3 flight controller. This work aimed to compare the accuracy of canopy temperature and to evaluate the performance of thermography. The extracted CT results were closely related to ground truth CT with the value of (R2) 0.9297 and correlation (r) 0.97702, respectively. The calculated results of CWSI showed a strong relation with gs under different irrigation levels 90%–100%, 75%, 60%, and 50% of the field capacity. The relationship of each time of day was substantial with (R2) 0.90, 0.75, and 0.86, respectively. The Correlation coefficients (R2) of CT, stomatal conductance, and SD were compared and found to be 0.755, 0.67, and 0.695, respectively. Results stated that this approach estimates the most reliable temperature around 35 °C to 40 °C. This study demonstrates the different temperature based spectral indices and provides accurate, rapid, and reliable canopy temperature quantification. © 2021 Elsevier B.V.